# SEO Split Into Two Jobs: What to Do in 2026 | Dexity

> SEO now has two targets: classic rankings and AI answer citations. See how to optimize for both when Google still drives ~95% of clicks.

Upskilling Reality

# SEO Didn't Die — It Split Into Two Jobs You Now Have to Do at Once

In 2026 SEO is a craft with two targets: ranking in classic search, and getting cited inside AI answers. The stakes are lopsided — AI search drives only ~1–2% of referral traffic for most sites, while Google still delivers ~95% of revenue-generating organic clicks (Aleyda Solis). But the AI slice is compounding fast: AI referral traffic grew 357% year over year. Here's the data, what an AI SEO engine actually is, and how to build one with Claude in a single sitting.

28 July 2026

7 min read

Author

[Shashi Bhushan Kumar](https://www.linkedin.com/in/shashi-bhushan-coder/)

Director - Product Management, Dexity

![An SEO specialist works two laptops at once beneath two wall screens, one showing a number one search ranking and the other an AI answer citing the site.](/assets/intel-ai-seo-hero-66f84d.webp)

On this page

1.  [What happened to SEO in the AI-search era?](#what-happened-to-seo-in-the-ai-search-era)
2.  [How much traffic does AI search drive vs. Google?](#how-much-traffic-does-ai-search-drive-vs-google)
3.  [What is an AI SEO engine?](#what-is-an-ai-seo-engine)
4.  [How do you build an AI SEO engine with Claude?](#how-do-you-build-an-ai-seo-engine-with-claude)
5.  [Build it live](#build-it-live)

Key facts

-   Google still drives **~95%** of revenue-generating organic clicks for most sites, according to Aleyda Solis.
-   AI search drives only **~1–2%** of referral traffic for most sites today, according to Aleyda Solis.
-   AI referral traffic grew **357%** year over year, per TechCrunch / Similarweb.
-   **65%** of the URLs ChatGPT cites sit two-to-three folders deep on a site, per Similarweb.
-   **~58.8%** of AI referral traffic lands on homepages rather than the deeper page that was cited, per Similarweb.

## What happened to SEO in the AI-search era?

Every few years someone declares SEO dead. In 2026 the obituary is wrong again — but the job genuinely changed shape. SEO is no longer one target. It's two. Google still drives ~95% of revenue-generating organic clicks, while AI search sends only ~1–2% of referral traffic today — yet that AI slice grew 357% year over year.

The first target is the one you already know: rank in classic search, where Google still sends the overwhelming majority of the clicks that turn into revenue. The second is new: get _cited_ inside AI answers — the paragraph ChatGPT or an AI overview quotes when someone asks your buyer's question and never scrolls to a blue link.

Both matter, and they don't compete. The mistake is treating this as a migration — abandoning what ranks to chase what's shiny. It isn't a migration. It's an addition. The craft now is ranking _and_ getting quoted, from the same content, on purpose.

## How much traffic does AI search drive vs. Google?

The numbers make the case for doing both — and for keeping your priorities straight about which one pays the bills today.

| Signal | Figure | Source |
| --- | --- | --- |
| Share of referral traffic AI search drives for most sites today | **~1–2%** | [Aleyda Solis](https://www.aleydasolis.com/en/) |
| Share of revenue-generating organic traffic Google still drives | **~95%** | [Aleyda Solis](https://www.aleydasolis.com/en/) |
| YoY growth in AI referral traffic | **357%** | [TechCrunch / Similarweb](https://www.similarweb.com/corp/reports/2026-generative-ai-landscape/) |
| URLs cited by ChatGPT that sit two-to-three folders deep | **65%** | [Similarweb](https://www.similarweb.com/corp/reports/2026-generative-ai-landscape/) |
| AI referral traffic that lands on homepages | **~58.8%** | [Similarweb](https://www.similarweb.com/corp/reports/2026-generative-ai-landscape/) |

Read those last two rows together and a strange gap appears: the page an AI _cites_ and the page that actually _gets the visit_ are frequently not the same. AI quotes your deep, specific article — then the human who follows through lands on your homepage. If you only optimize what ranks, you miss the citation. If you only chase citations, you starve the channel that pays.

The channel that gets cited and the channel that converts are drifting apart. An AI SEO engine's job is to close that gap on purpose — make the deep page quotable _and_ make the path from citation to conversion obvious.

## What is an AI SEO engine?

Strip away the tooling and it's a repeatable system with four moving parts:

1.  **Intent mapping.** Start from the questions your buyer actually types and asks — clustered by intent, not by keyword volume alone. Which clusters are worth ranking for? Which are worth being _quoted_ on? Rank them into a priority list before you write a word.
2.  **On-page briefs and drafts in your voice.** For each priority cluster, a brief that specifies the angle, the structure, and the questions the page must answer — then a draft written in your voice, not a generic AI mush that reads like every other page.
3.  **AEO — get cited by AI answers.** The same content, structured so an answer engine can lift a clean, self-contained passage: direct answers up top, clear definitions, FAQ blocks, sensible headings. Citable and rankable are not opposites; done well they're the same edit.
4.  **A cadence.** Not a one-time audit — a recurring loop. Publish, watch where you rank and where you're cited, find the next gap, repeat. SEO always rewarded consistency; the AI layer doesn't change that.

None of this requires becoming an engineer. It requires knowing what to prioritize, what to write, and how to keep score across both surfaces.

## How do you build an AI SEO engine with Claude?

You can stand up the whole engine with Claude in one sitting. The loop:

1.  **Cluster intent into a priority list.** Feed Claude your topic area and buyer questions; have it group them into intent clusters and rank them by a mix of business value and how winnable each is.
2.  **Generate an on-page brief.** For the top cluster, have Claude draft a brief — target intent, the sub-questions to answer, the structure, and the internal links — so the writing has a spec before it starts.
3.  **Draft in your voice, structured to rank.** Give Claude two or three samples of your existing writing and have it draft against the brief in your tone — with the heading hierarchy, internal links, and depth classic search rewards.
4.  **Structure the content to be citable by AI answers.** Have Claude add a clean direct-answer passage, a short definition, and an FAQ block — self-contained chunks an answer engine can quote without stitching your whole page together.
5.  **Monitor both surfaces.** Set a recurring check: where you rank in classic search, and where (and whether) you're cited across AI answers for your priority questions. The gaps become next week's briefs.
6.  **Repeat on a cadence.** One page proves the loop. The engine is running the loop every week without starting from scratch.

"In your voice" is doing real work here. Generic AI drafts are easy to spot and easy to ignore — by readers and, increasingly, by the systems ranking and citing them. The samples you feed Claude are what keep the output yours.

Don't torch your Google SEO to chase AI. With Google still driving ~95% of revenue-generating organic traffic and AI search at ~1–2%, gutting what ranks to go all-in on citations is trading a dollar for a penny. The move is additive: keep what ranks, and _also_ make it quotable. You're optimizing for where the 357% is heading — not abandoning where the revenue is.

## Build it live

Reading about the workflow and running it are different things. In Dexity's **AI for SEO workshop**, you build your AI SEO engine live in 90 minutes — guided by a practicing SEO who ranks content with AI. You cluster real intent, generate a brief, draft a page in your voice built to rank, structure it to be cited by AI answers, and set up the monitoring loop — walking away with it working, not with notes.

The people who own SEO in 2026 aren't the ones who can name the acronyms. They're the ones who can show the artifact: a page that ranks _and_ gets quoted, produced by a loop they run every week. That artifact is exactly what the workshop leaves you holding.

Join the AI for SEO workshop

## Frequently asked questions

Is SEO dead in 2026?

No. The job changed shape rather than ended. SEO is now two targets at once: rank in classic search, where Google still drives ~95% of revenue-generating organic traffic, and get cited inside AI answers, which is still a small but fast-growing slice.

What is an AI SEO engine?

It is a repeatable system with four moving parts: intent mapping (cluster buyer questions into a priority list), on-page briefs and drafts written in your voice, AEO so answer engines can lift a clean passage, and a recurring cadence to publish, check where you rank and get cited, and repeat.

How much traffic does AI search actually drive compared to Google?

AI search drives only ~1-2% of referral traffic for most sites today, while Google still delivers ~95% of revenue-generating organic clicks. The AI slice is compounding fast, though: AI referral traffic grew 357% year over year.

What does it take to get cited by AI answers (AEO)?

Structure the same content so an answer engine can lift a clean, self-contained passage: a direct-answer passage up top, clear definitions, an FAQ block, and sensible headings. Citable and rankable are not opposites; done well they are the same edit.

Should I move budget away from Google SEO to chase AI search?

No. With Google still driving ~95% of revenue-generating organic traffic and AI search at ~1-2%, gutting what ranks to go all-in on citations trades a dollar for a penny. The move is additive: keep what ranks, and also make it quotable.

Can I build an AI SEO engine with Claude?

Yes, you can stand up the whole loop with Claude in one sitting: cluster intent into a priority list, generate an on-page brief, draft in your voice structured to rank, add citable chunks for AI answers, monitor both surfaces, then repeat on a cadence. Feeding Claude two or three samples of your existing writing is what keeps the output in your voice.

* * *

_Sources: [Aleyda Solis](https://www.aleydasolis.com/en/) on AI-search vs. Google's share of revenue-generating organic traffic; [Similarweb — 2026 Generative AI Landscape Report](https://www.similarweb.com/corp/reports/2026-generative-ai-landscape/) (357% YoY AI referral growth via TechCrunch; citation-depth and homepage-landing figures via Aleyda Solis / Similarweb)._

## More in Upskilling Reality

[All Intel →](/intel)

[Upskilling Reality

### How to Set Up Conversion Tracking in Google Tag Manager (GA4, Google Ads, Meta) — and Automate It With AI in 2026

There are two ways to set up conversion tracking in Google Tag Manager in 2026. The manual way: build a GA4 event tag and mark the event as a Key Event; add a Google Ads conversion tag, relying on a site-wide Google Tag to store the GCLID (a standalone Conversion Linker is only needed on legacy containers without one); and install the Meta Pixel base with standard events (plus the Conversions API for server-side, deduplicated by a shared Event ID). The faster way: connect Claude to your container through a real GTM MCP server — like Stape's hosted google-tag-manager-mcp-server, which runs via npx mcp-remote at https://gtm-mcp.stape.ai/mcp on Node.js v18+ with Google OAuth — then run an audit, describe the tags you want, stage them in a new container version, and review before you publish. This guide gives the exact manual steps per platform, a copy-paste MCP config, a comparison of real GTM MCP servers, and the verification checklist for both paths — with the human review-before-publish gate kept non-negotiable.

1 August 2026 · 12 min read](/intel/gtm-tracking-with-ai) [Upskilling Reality

### AEO Is a Real Job Now — and the Fastest-Growing Skill in Marketing

Answer Engine Optimization (getting cited inside ChatGPT, Perplexity and Google AI answers) has stopped being a buzzword and become a hiring line item. AI referral traffic grew 357% year over year; 94% of CMOs are increasing AEO investment; and companies from Stripe to HubSpot to Anthropic are posting dedicated AEO roles at $75K–$210K. As Kaleigh Moore puts it, the function 'has separated from SEO the same way content marketing separated from copywriting a decade ago.' Here's the data, what the skill actually is, and how to build it in one sitting.

31 July 2026 · 8 min read](/intel/aeo-is-a-job-now) [Upskilling Reality

### AI Evals in Production: The Error-Analysis-First Playbook (2026)

Evals — not model choice, not prompt cleverness — decide whether AI features work in production. In Dexity's analysis of live job descriptions, evals now appear in 56% of AI-engineer and 32% of product-manager postings, up from near-zero two years ago. Here's the error-analysis-first method teams use to ship AI they can measure instead of hope for.

31 July 2026 · 8 min read](/intel/streamlining-ai-evaluation-in-production-with-data-collection-and-error-analysis)

Canonical: https://stg.dexity.com/intel/ai-seo-2026